Authentication System, Authentication Method, And, Non-Transitory Computer-Readable Information Recording Medium for Recording Program

1. An authentication system comprising: a storage configured to store, for each of a plurality of user IDs, a voiceprint of a user having authority pertaining to the user ID in association with a recorded sound including speech spoken previously by the user having the authority pertaining to the user ID; a specifier configured to (i) acquire an identifying sound including identifying speech spoken by a user attempting to receive authentication, (ii) recognize from the identifying sound by voice recognition a phrase read by the user, and (iii) specify a user ID allocated the recognized phrase as an identifying phrase and for which the identifying sound matches the voiceprint stored in association with the user ID; an outputter configured to output a masking sound by superimposing and outputting a part of all of the plurality of recorded sounds stored in association with the specified user ID; an acquirer configured to acquire a first sound that includes (i) a key-reading speech of the user reading an authentication key and (ii) the outputted masking sound; a remover configured to acquire a second sound by removing the masking sound from the acquired first sound; and a determiner configured to determine whether the user has the authority pertaining to the specified user ID based on the acquired second sound, wherein the determiner uses, as necessary conditions for determining that the user has the authority pertaining to the specified user ID, (i) recognizing from the second sound a word expression read by the user, and the recognized word expression being valid as the authentication key for the specified user ID, and (ii) the second sound matching the voiceprint stored in association with the specified user ID.

2. The authentication system according to claim 1 , further comprising: an updater configured to, upon determination that the user has the authority pertaining to the specified user ID, append or save by overwriting to the storage, as the recorded sound and in association with the specified user ID, (i) the second sound or (ii) a result of dividing the second sound in accordance with units constituting the authentication key.

3. The authentication system according to claim 1 , wherein the authentication system: provides to the user, together with the outputting of the masking sound, a random table in which one of a plurality of random values is embedded in each element of elements of the random table, a selection rule indicating an order of selection of positions of the elements included in the table being associated beforehand with the user ID for each of the user IDs; extracts, from the provided random table, a value string arranging the values embedded in the elements selected by the selection rule associated with the specified user ID; and takes the recognized word expression to be valid as the authentication key pertaining to the specified user ID when the recognized word expression and the extracted value string match each other.

4. The authentication system according to claim 1 , wherein prior to specifying the user ID, the authentication system: acquires an identifying sound that includes an identifying speech spoken by the user attempting to receive authentication; uses voice recognition to recognize from the identifying sound a phrase read by the user; and specifies a user ID that is a user ID allocated the recognized phrase as the identifying phrase and for which the identifying sound matches the voiceprint stored in association with the user ID.

5. The authentication system according to claim 1 , wherein for each of the user IDs, the authentication system allocates an identifying phrase that is different from the identifying phrase of the other user IDs.

6. The authentication system according to claim 1 , wherein the authentication system: for each user having the authority pertaining to the user ID of the plurality of user IDs, allocates a plurality of identifying phrases; and when a plurality of the user IDs is specified, until just a single user ID is specified, repeatedly (i) prompts the user to read a new identifying phrase, (ii) acquires a new identifying sound that includes a new identifying speech spoken by the user, and (iii) specifies the user ID based on the new identifying sound.

7. An authentication method, by an authentication system that, for each of a plurality of user IDs, stores a voiceprint of a user having authority pertaining to the user ID in association with a plurality of recorded sounds including a plurality of speeches previously spoken by the user having the authority pertaining to the user ID, the method comprising: acquiring an identifying sound including an identifying speech spoken by a user attempting to receive authentication, recognizing from the identifying sound by voice recognition a phrase read by the user, and specifying a user ID for which the recognized phrase is allocated as an identifying phrase and for which the identifying sound matches the voiceprint stored in association with the user ID; outputting a masking sound by superimposing and outputting a part of all of the plurality of recorded sounds stored in association with the specified user ID; acquiring a first sound including (i) a key-reading speech of the user reading the authentication key and (ii) the outputted masking sound; acquiring a second sound by removing the masking sound from the acquired first sound; and determining whether the user has the authority pertaining to the specified user ID based on the acquired second sound, wherein necessary conditions for determining that the user has authority pertaining to the specified user ID are: (i) a word expression read by the user is recognized from the second sound, and the recognized word expression is valid as the authentication key for the specified user ID, and (ii) the second sound matches the voiceprint stored in association with the specified user ID.

8. A non-transitory computer-readable information recording medium for recording a program for causing a computer or a computer group, for each of a plurality of user IDs, storing a voiceprint of a user having authority pertaining to the user ID in association with a plurality of recorded sounds including a plurality of speeches previously spoken by the user having the authority pertaining to the user ID, to execute processing comprising: acquiring an identifying sound including an identifying speech spoken by a user attempting to receive authentication, recognizing from the identifying sound by voice recognition a phrase read by the user, and specifying a user ID for which the recognized phrase is allocated as an identifying phrase and for which the identifying sound matches the voiceprint stored in association with the user ID; outputting a masking sound by superimposing and outputting a part of all of the recorded sounds stored in association with the specified user ID; acquiring a first sound including (i) a key-reading speech of the user reading the authentication key and (ii) the outputted masking sound; acquiring a second sound by removing the masking sound from the acquired first sound; and determining whether the user has the authority pertaining to the specified user ID based on the acquired second sound, wherein necessary conditions for determining that the user has authority pertaining to the specified user ID are: (i) a word expression read by the user is recognized from the second sound, and the recognized word expression is valid as the authentication key for the specified user ID, and (ii) the second sound matches the voiceprint stored in association with the specified user ID.